Protein covalent immobilization via its scarce thiol versus abundant amine groups: Effect on orientation, cell...

Quantity, orientation, conformation and covalent linkage of naturally cell adhesive proteins adsorbed or covalently linked to a surface, are known to influence the preservation of their subsequent long term cell adhesion properties and bioactivity. In the present work, we explore two different strategies for the covalent linking of plasma fibronectin (pFN) - used as a cell adhesive model protein, onto a polystyrene (PS) surface. One is aimed at tethering the protein to the surface in a semi-oriented fashion (via one of the 4 free thiol reactive groups on the protein) with a heterofunctional coupling agent (SSMPB method). The other aims to immobilize the protein in a more random fashion by reaction between the abundant pendant primary amine bearing amino acids of the pFN and activated carboxylic surface functions obtained after glutaric anhydride surface treatment (GA method). The overall goal will be to verify the hypothesis of a correlation between covalent immobilization of a model cell adhesive protein to a PS surface in a semi-oriented configuration (versus randomly oriented) with promotion of enhanced exposure of the protein's cell binding domain. This in turn would lead to enhanced cell adhesion. Ideally the goal is to elaborate substrates exhibiting a long term stable protein monolayer with preserved cell adhesive properties and bioactivity for biomaterial and/or cell adhesion commercial plate applications. However, the initial restrictive objective of this paper is to first quantitatively and qualitatively investigate the reversibly (merely adsorbed) versus covalently irreversibly bound protein to the surface after the immobilization procedure. Although immobilized surface amounts were similar (close to the monolayer range) for all immobilization approaches, covalent grafting showed improved retention and stronger "tethering" of the pFN protein to the surface (roughly 40%) after SDS rinsing compared to that for mere adsorption (0%) suggesting an added value to the covalent grafting immobilization methods. However no differences in exposure of the cell binding domains were observed (ELISA results) before SDS rinsing, suggesting that pFN protein grafting to the surface is initially kinetically driven be a stochastic random adsorption phenomenon. Covalent grafting acts in the final stage as a process that simply tethers and stabilizes (or freezes) the initial conformation/orientation of the adsorbed protein on the surface. In addition covalent linkage via the SSMPB approach is likely favored by surface-induce exposure of one of the normally hidden free thiol group pair, thus optimizing covalent linkage to the surface. However after SDS rinsing, this "tethering"/"freezing" effect was significantly more prominent for the GA grafting approach (due to greater number of potential covalent links between the protein and the surface) compared to that for the SSMPB approach. This hypothesis was buttressed by the improved resistance to denaturation (smaller conformational lability) for the GA compared to the SMPB approach and improved exposure of the cell binding domain for the former (>50%) even after SDS rinsing. These results are promising in that they suggest covalent tethering of fibronectin to PS substrate in a monolayer range, with significantly improved irreversible protein surface bonding via both approaches (compared to that for mere adsorption). The latter are likely applicable to a wide range of proteins.